Tumors particularly, particularly clinical and experimental melanomas, express peptides that can be presented on MHC class I products to cytolytic T lymphocytes [CTLs]. The example to be studied here is the mouse B16 melanoma, for which gp175 and TRP-2 peptides have been defined. Yet these peptide-specific T cells are not expanded in tumor bearing mice. We hypothesize that tumors are not presented by the dendritic cell [DC] system that normally initiates immunity, and therefore, that targeting of melanoma to DCs should elicit tumor immunity and resistance. Under the appropriate developmental conditions, DCs efficiently form MHC-peptide complexes, express high levels of T cell co-stimulatory molecules, and migrate to lymphoid tissues to initiate immunity. Previously, potent and specialized antigen presenting DCs have not been available in sufficient numbers, nor has it been evident how to load DCs with cellular antigens. These obstacles are being overcome. Now it is possible to generate large numbers of cells from mouse marrow precursors, and these process peptides from other apoptotic and necrotic cells. Therefore our aims are to 1] charge DCs with dying B16 melanoma ex vivo and use these to optimize the induction of tumor immunity in vivo; 2] target dying B16 melanoma cells to DCs in situ to induce immunity; 3] evaluate if DCs can process a helper epitope within B16 tumors that are growing in situ; and 4] analyze subsets of DCs for their capacity to elicit melanoma immunity.